266 Manometric Observations in the Electric Arc. [Jan. 19 ? 



arc is a matter of convenience, as in this condition the manometers 

 are least affected by air currents, but the same general action may be 

 observed by the use of vertical poles. A thin carbon spatula placed in 

 front of the end of the positive pole at once lowered the positive pres- 

 sure of the manometer. This experiment does not prove much, as the 

 carbon diaphragm causes a noisy and unsteady arc until a minute hole 

 is pierced by the current. It is probable that the diminution of pressure 

 may be due to the instability of the arc. In the same way the action of 

 the magnet on the arc causes an instant reduction of the positive pres- 

 sure by withdrawing the arc from completely covering the end of the 

 carbon tube, so that this action must be regarded as an indirect one. 

 A small carbon tube connected with a manometer was inserted into 

 the arc, passing between two solid carbons, so as to take a section at 

 right angles to the passage of the current. In this condition the arc 

 is apt to be irregular in shape, and to pass rather between three poles 

 than two, but the average records point to an increase of pressure in this 

 position also. When the negative carbon tube is about 1 millim. in 

 diameter, and the point sharp and the tube short, so as to diminish 

 any air friction, the negative pole manometer seems also to give a 

 positive pressure. The intermittent Siemens arc shows an increase of 

 pressure at both poles. 



It appears from the above experiments that the interior of the 

 gaseous envelope of the electric arc always shows a fixed permanent 

 pressure, amounting to about a millimetre of water above that of the 

 surrounding atmosphere. This looks as if the well-defined boundary 

 of the heated gases acted as if it had a small surface tension. This 

 pressure may be due to various causes : motion of the gas par- 

 ticles under the conditions, transit of material from pole to pole, or 

 a succession of disruptive discharges ; and a more elaborate investiga- 

 tion will have to be made before the origin of the excess of pressure 

 can be clearly defined. 



